GCD AI RADAR
incidents/AIR-2026-1393

Model-as-a-Trojan: Self-Activating Covert DNS Exfiltration in LLM Weights

Attack on AI1 Research / proof of conceptSupply chainAgent hijackData leak

Researchers demonstrated that fine-tuning open-weight models can embed covert DNS exfiltration behaviors directly into model parameters. The attack triggers malicious shell execution and data leakage during agent tool use.

Details

type
Attack on AI
An AI model, agent or AI-powered product was the thing exploited, abused or leaked.
categories
  • Supply chainprimary
    Malicious or backdoored models, poisoned packages, unsafe deserialization and other compromised AI dependencies.
  • Agent hijack
    An AI agent is tricked into unauthorized actions: tool misuse, data exfiltration through the agent, or acting against its principal.
  • Data leak
    Training data, chat logs, system prompts, model weights or other AI-system data exposed.
owasp llm top 10
mitre atlas
Both rows are derived from the type and categories (OWASP 2025 edition, ATLAS release 2026.09), not classified per record. How
severity
1 · Research / proof of concept
Demonstrated by researchers; no known real-world exploitation.
targets
Open-weight model
products
open-weight LLM
incident date
19 Oct 2026
first reported
19 Oct 2026
origin
News & research feeds
classifier confidence
0.95 · published at or above 0.80; see the methodology
tags
model-trojan · dns-exfiltration · supply-chain · weight-poisoning
record
created 2026-10-19 05:07Z · updated 2026-10-19 05:07Z · json

Sources · 1

  1. 2026-10-19
    Model-as-a-Trojan: Self-Activating Covert DNS Exfiltration Baked Into the Weights of an LLM Agent
    arXiv · arxiv.org/abs/2610.22578

Titles and links belong to their publishers. The summary above is written by GCD AI RADAR in its own words; data is CC BY-SA 4.0.

Related incidents